U.S. patent application number 15/526657 was filed with the patent office on 2018-03-01 for distributed optical sensing using compressive sampling.
The applicant listed for this patent is Halliburton Energy Services, Inc.. Invention is credited to David Andrew Barfoot, Safyan Gopal Bhongale, Andreas Ellmaufhaler, Christopher Lee Stokley.
Application Number | 20180058197 15/526657 |
Document ID | / |
Family ID | 59225128 |
Filed Date | 2018-03-01 |
United States Patent
Application |
20180058197 |
Kind Code |
A1 |
Barfoot; David Andrew ; et
al. |
March 1, 2018 |
DISTRIBUTED OPTICAL SENSING USING COMPRESSIVE SAMPLING
Abstract
Distributed optical sensing systems utilize compressive sensing
techniques to determine parameters sensed by a waveguide. The
system generates light that is sent along a sensing waveguide,
thereby producing backscattered light. A compressive sampling
filter forms part of the system, and is used to selectively block
portions of the generated light or the backscattered light. The
backscattered light is received by a receiver and used to determine
one or more parameters.
Inventors: |
Barfoot; David Andrew;
(Houston, TX) ; Bhongale; Safyan Gopal; (Cypress,
TX) ; Stokley; Christopher Lee; (Houston, TX)
; Ellmaufhaler; Andreas; (Rio de Janer o, BR) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Halliburton Energy Services, Inc. |
Houston |
TX |
US |
|
|
Family ID: |
59225128 |
Appl. No.: |
15/526657 |
Filed: |
December 28, 2015 |
PCT Filed: |
December 28, 2015 |
PCT NO: |
PCT/US2015/067598 |
371 Date: |
May 12, 2017 |
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01V 8/22 20130101; G01D
5/35358 20130101; E21B 47/135 20200501; G01N 21/03 20130101; G01N
2021/0364 20130101; E21B 47/113 20200501; G01V 8/10 20130101; G01V
8/24 20130101; G01V 8/02 20130101 |
International
Class: |
E21B 47/10 20060101
E21B047/10; E21B 47/12 20060101 E21B047/12; G01V 8/24 20060101
G01V008/24; G01D 5/353 20060101 G01D005/353; G01N 21/03 20060101
G01N021/03 |
Claims
1. A distributed optical sensing system, comprising: a light source
to generate light; a waveguide positioned to optically interact
with the light to produce backscattered light; a compressive
sampling filter positioned to selectively block portions of the
light or the backscattered light based upon a compressive sampling
technique; and an optical receiver to receive the backscattered
light.
2. A system as defined in claim 1, wherein the compressive sampling
filter is an optical switch or optical shutter.
3. A system as defined in claim 1, wherein the compressive sampling
filter is optically coupled between the waveguide and the optical
receiver to selectively block portions of the backscattered
light.
4. A system as defined in claim 1, wherein the optical receiver has
a bandwidth of 10 MHz or less.
5. A system as defined in claim 1, wherein the compressive sampling
filter is optically coupled between a local oscillator and a mixer
to selectively block portions of the backscattered light.
6. A system as defined in claim 5, wherein the mixer is a 90 degree
hybrid mixer.
7. A system as defined in claim 5, wherein the optical receiver is
a balanced optical receiver.
8. A system as defined in claim 1, wherein the compressive sampling
filter is optically coupled between the light source and the
waveguide to selectively block portions of the light.
9. A system as defined in claim 8, wherein the light source
generates a light pulse having width of 1 microsecond or
longer.
10. A system as defined in claim 1, further comprising processing
circuitry communicably coupled to the receiver to receive the
backscattered light and thereby determine a parameter being sensed
by the waveguide.
11. A system as defined in claim 1, wherein the waveguide is
positioned along a wellbore.
12. A system as defined in claim 1, wherein the waveguide is a
fiber optic cable.
13. A distributed optical sensing method, comprising: generating
light that optically interacts with a waveguide to produce
backscattered light; selectively blocking portions of the light or
the backscattered light using a compressive sampling filter; and
analyzing the backscattered light to thereby determine a parameter
being sensed by the waveguide.
14. A method as defined in claim 13, wherein selectively blocking
the backscattered light comprises optically interacting the
backscattered light with the compressive sampling filter.
15. A method as defined in claim 13, wherein selectively blocking
the backscattered light is comprises using an optical switch or
optical shutter to selectively block the backscattered light.
16. A method as defined in claim 13, wherein selectively blocking
the backscattered light comprises: optically interacting a local
oscillator light with the compressive sampling filter to produce a
reference light; optically interacting the reference and
backscattered light with a mixer to produce mixed light; and
optically interacting the mixed light with the receiver to thereby
generate the backscattered light.
17. A method as defined in claim 16, wherein optically interacting
the backscattered light further comprises phase shifting the
reference light 90 degrees relative to the backscattered light.
18. A method as defined in claim 13, wherein selectively blocking
the backscattered light comprises optically interacting the light
with the compressive sampling filter to produce a compressive
sampling pulse having blocked regions.
19. A method as defined in claim 18, wherein generating the light
comprises generating a light having a pulse width of 1 microsecond
or longer.
20. A method as defined in claim 13, wherein determining the
parameter comprises determining a parameter along a wellbore.
21. A distributed optical sensing method, comprising: interrogating
a sensing waveguide using an interrogation signal; receiving
backscattered signals from selected regions of the sensing
waveguide based upon a compressive sampling technique; and using
the backscattered signals, determining a parameter being sensed by
the sensing waveguide.
22. A method as defined in claim 21, wherein receiving
backscattered signals comprises selectively blocking portions of
the backscattered signals using a compressive sampling filter
positioned in-line with the sensing waveguide.
23. A method as defined in claim 21, wherein receiving
backscattered signals comprises selectively blocking portions of a
local oscillator light.
24. A method as defined in claim 21, wherein receiving
backscattered signals comprises mixing the backscattered light with
a reference signal.
25. A method as defined in claim 21, wherein receiving
backscattered signals comprises selectively blocking portions of
the interrogation signal.
26. A method as defined in claim 21, wherein a wellbore parameter
is determined.
Description
FIELD OF THE DISCLOSURE
[0001] The present disclosure relates generally to optical sensing
and, more specifically, to an optical sensing using compressive
interrogation methods.
BACKGROUND
[0002] In distributed fiber optic sensing, an unmodified fiber
optic waveguide is used as a sensor. There are many ways to
interrogate a distributed fiber optic sensor, but all of these
methods require sending optical energy into the fiber to produce a
backscattering of the light which is used to measure a physical
property in proximity to the fiber, such as, temperature,
vibration, static or dynamic strain, chemical concentration, or
pressure. Examples of commercially established methods include
distributed temperature sensing ("DTS"), distributed acoustic
sensing ("DAS"), and distributed strain sensing ("DSS").
[0003] In such systems, it is desirable to spatially divide the
fiber optic cable, which may be many kilometers long, into discrete
sensing regions so that the cable is transformed into a dense array
of sensors. These sensors are not placed in or attached to the
fiber, but instead are created by the way the fiber is
interrogated. Ideally, the dense array of sensors will have spacing
between the sensors down to the level of a few meters or less if
possible, so as to achieve a very fine spatial resolution. This is
of particular importance when sensing along the length of a
wellbore, where features of interest may be very localized and in
close proximity to areas with different properties. For example,
wellbore features like perforation clusters, packers, and
production zones may need to be spatially separated in any
effective measurement.
[0004] In distributed sensing, when a light interrogation signal is
sent into the fiber, as the light is travelling down the fiber, a
continuous backscatter signal is generated. The backscatter signal
of interest may consist of one or a combination of Rayleigh,
Brillouin, or Raman backscatter. In order to provide an array of
sensors spaced closely along the fiber, a method of multiplexing,
or dividing up the backscatter response from these sensing regions
must be utilized. Well known methods for optical spatial
multiplexing include time domain multiplexing, frequency domain
multiplexing, and code-division multiplexing.
[0005] However, the data storage and processing requirements of
conventional systems are disadvantageous. In order to achieve a
desired spatial resolution, the optical receiver must detect and
sample the backscattered signal at a high speed. As a result, high
speed and bandwidth system components are necessary. However, as
the bandwidth increases, the performance of the receiver will be
degraded by a proportional amount, thereby adversely affecting the
integrity of the sensed parameters.
BRIEF DESCRIPTION OF THE DRAWINGS
[0006] FIG. 1 is a block diagram of a distributed optical sensing
system using an optical switch or shutter as a compressive sampling
filter, according to certain illustrative embodiments of the
present disclosure;
[0007] FIG. 2 is a block diagram of a distributed optical sensing
system where the backscattered light is mixed with a local
oscillator light, according to certain illustrative embodiments of
the present disclosure;
[0008] FIG. 3 illustrates a distributed optical sensing system in
which compressive sampling is applied to the light pulse, according
to certain illustrative alternative embodiments of the present is
disclosure;
[0009] FIG. 4 is a diagram showing three different compressive
sampling pulses being sent through a sensing waveguide; and
[0010] FIG. 5 is a schematic illustration of a distributed optical
sensing system extending along a wellbore, according to alternative
illustrative embodiments of the present disclosure.
DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS
[0011] Illustrative embodiments and related methods of the present
disclosure are described below as they might be employed in a
distributed optical sensing system using compressive sensing. In
the interest of clarity, not all features of an actual
implementation or methodology are described in this specification.
It will of course be appreciated that in the development of any
such actual embodiment, numerous implementation-specific decisions
must be made to achieve the developers' specific goals, such as
compliance with system-related and business-related constraints,
which will vary from one implementation to another. Moreover, it
will be appreciated that such a development effort might be complex
and time-consuming, but would nevertheless be a routine undertaking
for those of ordinary skill in the art having the benefit of this
disclosure. Further aspects and advantages of the various
embodiments and related methodologies of the disclosure will become
apparent from consideration of the following description and
drawings.
[0012] As described herein, illustrative embodiments of the present
disclosure are directed distributed optical sensing systems which
utilize compressive sensing techniques. In a generalized
embodiment, the sensing system includes a light source to generate
light that is sent along a sensing waveguide, thereby producing
backscattered light. A compressive sampling filter forms part of
the system, and is used in a variety of ways to selectively block
portions of the light generated by the light source or the
backscattered light. The backscattered light is then received by
one or more optical receivers and used by processing circuitry to
determine a parameter being sensed by the waveguide. Accordingly,
as described herein, a system is provided that utilizes compressive
sensing to enhance the performance of distributed fiber optic
sensors, while simultaneously reducing data storage and bandwidth
requirements by a significant amount. Alternatively, embodiments of
the present disclosure may be considered as providing data
compression embedded in the sensing device itself. Furthermore,
such advantages also is translate into higher resolution, when
compared to conventional systems, without increasing the sampling
rate.
[0013] As previously mentioned, there are various methods for
optical spatial multiplexing. Although applicable to others, the
illustrative embodiments and methods described herein will focus on
time domain multiplexing. In an optical time-domain reflectometer,
a short duration (i.e., 10 ns) optical pulse is sent into a fiber
optic sensing cable or other waveguide. As the light pulse travels
down the length of the fiber, it generates a continuously
backscatter signal from a continuum of points along the fiber which
can be separated in time as it arrives back at the optical
interrogation system.
[0014] To achieve a desired spatial resolution, an optical receiver
will detect and sample the backscatter signal at a high speed. In
typical fiber optic systems, the light will travel at approximately
5 ns per meter in the fiber. Therefore, the two-way travel time of
the interrogation light to travel from the interrogation system,
arrive at a particular location, and subsequently for the
backscatter from that location to arrive back at the optical
interrogation unit is 10 ns per meter (5 ns/meter in each
direction). If the optical signal is sampled by the receiver every
10 ns, or at 100 MHz, the backscatter from each one meter section
of fiber will be sampled separately, leading to a dense array of
sensors with the center of each sensing region spaced one meter
apart. For example, with a 10 km long fiber optic cable, a 100 MHz
sample rate will provide 10,000 discrete sensors.
[0015] From this description, if it is desirable to increase the
spatial resolution of the sensor array by reducing the spacing
between the sensing regions, the optical backscatter signal must be
sampled at a faster rate. For example, to achieve a 10 cm spacing
between the sensing regions, the backscatter must be sampled at 1
GHz. For smaller spacing between sensing regions, a higher rate
would be needed.
[0016] In order to achieve a high spatial resolution in general,
both the optical receiver and analog-to-digital converter ("ADC")
must support a bandwidth compatible with the sampling rate required
for the desired spatial resolution. For example, in a typical
system with one meter spatial resolution, the optical receiver will
be designed to have at least a 100 MHz bandwidth. The ADC will be
required to sample at least twice this rate, or 200 Mega samples
per second ("MSPS"), based on the Nyquist criteria requiring
sampling at twice the frequency of the highest frequency component
of the signal to prevent aliasing which would translate into an
undesirable is noise added to the signal being measured.
[0017] While in theory it is possible to achieve very high spatial
resolution with the time domain multiplexing method, many
compromises must be made when increasing the bandwidth of optical
receivers and electronic components following the receiver,
including amplifiers and ADCs. A typical optical receiver will have
a gain (typically a trans-impedance gain), and saturation power
level (maximum optical power above which the receiver will not
function correctly), and background noise or self-noise (often
given in terms of Noise Equivalent Power or "NEP"). As the
bandwidth of the receiver is increased, many or all of these other
parameters will be degraded due to the required changes to the
device to achieve higher bandwidth.
[0018] Typically, the product of the gain and bandwidth is a
constant for any particular type of optical receiver. This is
referred to as the "gain-bandwidth product." Therefore, as the
bandwidth is increased, the gain must be reduced by a proportional
amount. Because the backscatter signals are very weak compared to
the interrogation signal, high gain for the optical receiver is
very important in most applications. In addition, low self-noise is
important. The NEP may also be degraded when the bandwidth of the
receiver is increased. Similar compromises are made with ADCs. For
example, to precisely convert an electrical voltage into a digital
number, the voltage must be converted into a series of bits. The
more bits that are available, the more precisely the voltage can be
digitized for later computations. If the number of bits available
is too low for the particular measurement, quantization noise
becomes a dominant source of noise. Typically, the number of bits
available for a commercial ADC is inversely proportional to the
maximum sample rate. For example, today many ADCs are available at
100 MSPS with 16-bits of resolution. At speeds over 1 GSPS, more
commonly 8-bit or 12-bit ADCs can be found, but not the 16-bit ADCs
that are desirable for many applications of fiber optic
sensing.
[0019] In view of the foregoing, the illustrative embodiments and
methods described herein use lower speed and bandwidth components
that provide increased performance in terms of gain, dynamic range,
and resolution, while simultaneously maintaining a desirable
spatial resolution for a particular application. To achieve this,
the present disclosure applies compressive signal sampling
techniques, also referred to as compressive sensing or compressed
sensing. One example of compressive sampling in use is with a
single pixel camera, such as described in K. Kelly, et. al., "An
Architecture for Compressive Imaging," 2006 IEEE international
Conference is on Image Processing, Oct. 8-11 2006, pg. 1273-1276.
The main idea of compressive sampling is, instead of measuring all
components of a signal separately (by high-speed sampling or by
using a sensor array as in a camera), many, but not all of the
components of a signal are combined into a single measurement.
These composite measurements are performed many times, where in
each case, the components (pixels, etc.) that are combined together
into the measurement are in most cases randomly chosen using a
pseudo-random spatial light filtering method.
[0020] In the single pixel camera sampling method, an image is
acquired using a single pixel camera. Each measurement includes the
combined signal components of what would normally be sensed at many
different pixel locations. An array of mirrors is modulated for
each measurement so that a pseudo-random combination of pixel
locations is combined together at a single optical sensor. By
taking many measurements, but far fewer than the total number of
image pixels in the final image, it is possible through compressive
sampling signal recovery methods, such as, for example, those
described in E. Candes, M. Wakin; "An Introduction to Compressive
Sampling," IEEE Signal Processing Magazine, March 2008 pg. 21-30
and K. Kelly, et. al., "An Architecture for Compressive Imaging,"
2006 IEEE international Conference on Image Processing, Oct. 8-11
2006, pg. 1273-1276, to reconstruct the full image signal from
these single pixel samples.
[0021] Turning to illustrative embodiments of the present
disclosure, in distributed fiber optic sensing, the spatial samples
of an image can be looked at as equivalent if the spatial domain is
transformed into the time domain. In other words, the spatial
separation of fiber sensors translates to time at a single
photodetector/receiver of an interrogation system. Replacing the
conventional high-bandwidth receiver system with a low-bandwidth
receiver system of the present disclosure would cause the signals
from many adjacent fiber sensors to be combined together into a
single measurement. This is similar to what happens with the single
pixel camera when an array of signals assigned to different pixels
is combined into a single pixel detector. In the case of the
single-pixel camera, micro-mirrors are used to allow only some of
the pixels to be received at the detector instead of all of them.
As will be described in detail below, the illustrative embodiments
herein utilize a compressive sampling filter to perform this same
function, thereby selectively blocking certain portions of
backscattered light.
[0022] FIG. 1 is a block diagram of a distributed optical sensing
system using an optical switch or shutter as a compressive sampling
filter, according to certain illustrative embodiments of the
present disclosure. In this example, distributed optical sensing
system 10 includes a light source 2 to generate a light used to
interrogate a waveguide 8. Light source 2 may be, for example, a
laser or other suitable source of electromagnetic radiation. A
modulator 4 is optically coupled in-line with light source 2 in
order to pulsate or otherwise modulate the generated light.
Modulator 4 is optically coupled to waveguide 8 via an optical
circulator 6 which facilitates bi-directional communication of
light signals. Waveguide 8 may be, for example, a fiber optic
cable. During operation, light source 2 generates light that is
modulated as desired by modulator 4 and sent down waveguide 8,
thereby creating backscattered light signals which are sent back up
waveguide 8 toward optical circulator 6 and the receiver components
of system 10.
[0023] Distributed optical sensing system 10 also includes a
compressive sampling filter 18 (e.g., a high-speed optical shutter
or switch) to perform the function of blocking the backscattered
signal from selective spatial locations of waveguide 8 (e.g., fiber
optic cable) while allowing other locations to pass to the
low-bandwidth receiver 12. In this embodiment, compressive sampling
filter 18 is optically coupled between waveguide 8 and receiver 12.
Receiver 12, because it is low bandwidth (e.g., 1 MHz, 100 KHz,
etc.), will provide a measurement in proportion with the average of
the combined spatial locations that are allowed by pass through
compressive sampling filter 18.
[0024] To achieve this, a compressive sampling spatial
randomization signal 16 is supplied to compressive sampling filter
18. A spatial randomization signal is a sequence of binary numbers
generated using a pseudo-random number generator, as will be
understood by those ordinarily skilled in the art having the
benefit of this disclosure. Spatial randomization signal 16 enables
compressive sampling filter 18 to perform pseudo-random compressive
sampling by driving, in this example, an ultrafast optical shutter
to open or close based on the timing of the backscattered light
from the particular regions of waveguide 8 that are desired to be
combined at receiver 12. An ultrafast optical shutter or switch may
be one that, for example, switches or opens/closes at times below 1
ns. After the backscattered light is received at receiver 12, it is
then transmitted to ADC 14 to digitize the backscattered light for
processing by processing circuitry on-board or remote from ADC 14
to thereby determine one or more parameters sensed by waveguide
8.
[0025] Going back to spatial randomization signal 16, as an
example, if a one meter spatial is resolution is desired, the
shutter of compressive sampling filter 18 may be modulated at a
speed of at least 100 MHz, for example, thus blocking the
backscatter light from pseudo-randomly selected one meter sections
of waveguide 8. The low bandwidth (i.e., 1 MHz, 100 kHz, etc.)
optical receiver 12 will provide an integration of the optical
energy from many locations at once. In certain illustrative
embodiments, optical receiver 12 has a bandwidth of 10 MHz or less.
Thus, instead of processing each individual backscattered light
energy (which requires higher bandwidth and processing capability),
a plurality of backscattered light energies are combined and
integrated (thus, providing allowances for lower bandwidth and
processing capability). Ultimately, this measurement of the
parameter being sensed by the corresponding portion of waveguide 8
may be of a significantly higher fidelity (SNR, gain, etc.) due to
the use of a much lower bandwidth receiver 12 (as compared to
conventional optical time domain reflector ("OTDR") methods).
[0026] In addition, these lower bandwidth components, both optical
receiver 12 and low speed ADC 14, are typically much lower in cost
and power consumption. ADC 14 can be lower bandwidth because
receiver 12 is low bandwidth. A low bandwidth ADC may be, for
example, a Texas Instruments.RTM. model ADS1675. This has benefits
for embedded applications that must run off of battery power or
cost sensitive applications. If it is desirable to have a spatial
resolution of 1/2 meter, the same low-bandwidth optical receiver
and ADCs can be used with the only change being that the optical
shutter is modulated at twice the rate. Therefore, the bandwidth
requirements of optical receiver 12 and ADC 14 discussed previously
are transferred to compressive sampling filter 18.
[0027] In this illustrative embodiment, many devices may service as
compressive sampling filter 18. For example, a semiconductor
optical amplifier ("SOA") can act as a fast optical switch with
switching times below 1 ns (1 GHz), for example, Thorlabs.TM. model
SOA1013SXS. Other examples include electro-optic devices, such as a
Lithium Niobate (LiNbO3) or Indium Phosphide (lnP) waveguide in a
Mach-Zehnder configuration, which can be used as an optical shutter
with a bandwidth above 10 GHz. Also, acousto-optic devices, like a
Bragg cell, can be used as a shutter by using the deflection
properties of the Bragg cell to redirect the light passing through
so that it doesn't line up with the exit fiber, thus acting as a
very fast shutter. Electro-absorption modulators, for example made
from lnP may also be used as a high-speed shutter. Faster optical
shutters using vanadium oxide (V02) may also be used which have the
potential to is switch at Terahertz speeds, as described in D.
Johnson, "Nanoscale Metamaterial Optical Switches Operate at
Terahertz Speeds," IEEE Spectrum, Mar. 14, 2014. By switching at
very high speeds (e.g., GHz) a spatial resolution may be provided
at 10 cm or below levels while using low cost and low bandwidth
optical receivers and electrical components. Moreover, in certain
illustrative embodiments, in order to achieve these low spatial
resolutions, it would also be necessary for the pulse width to be
no greater than the desired spatial resolution.
[0028] Moreover, although not shown, the interrogation and
receiving components of optical sensing system 10 are communicably
coupled to processing circuitry (and may be jointly referred to as
a "light detection unit"). The light detection unit may, among
other functions, control the operation of the light source, spatial
randomization signal, etc., as well as the processing of the
received backscattered light signals for determination of sensed
parameters. Therefore, the light detection units described herein
may include at least one processor, a non-transitory,
computer-readable storage, transceiver/network communication
module, optional I/O devices, and an optional display (e.g., user
interface), all interconnected via a system bus. The network
communication module may be any type of communication interface
such as a fiber optic interface and may communicate using a number
of different communication protocols. Software instructions
executable by the processor for implementing the compressive
sampling described herein may be stored in suitable storage or some
other computer-readable medium.
[0029] Moreover, those skilled in the art will appreciate that the
disclosure may be practiced with a variety of computer-system
configurations, including hand-held devices, multiprocessor
systems, microprocessor-based or programmable-consumer electronics,
minicomputers, mainframe computers, and the like. Any number of
computer-systems and computer networks are acceptable for use with
the present disclosure. The disclosure may be practiced in
distributed-computing environments where tasks are performed by
remote-processing devices that are linked through a communications
network. In a distributed-computing environment, program modules
may be located in both local and remote computer-storage media
including memory storage devices. The present disclosure may
therefore, be implemented in connection with various hardware,
software or a combination thereof in a computer system or other
processing system.
[0030] FIG. 2 is a block diagram of a distributed optical sensing
system where the backscattered is light is mixed with a local
oscillator light, according to certain illustrative embodiments of
the present disclosure. In the case of distributed acoustic
sensing, it is not necessary to position the compressive sampling
filter as described in relation to FIG. 1. FIG. 2 illustrates an
alternate embodiment for a distributed optical sensing system 20
whereby the backscattered light is mixed with a much more power
laser light (i.e., local oscillator light). FIG. 2 is similar to
the embodiment of FIG. 1, whereby like elements are identified
using like numerals. However, instead of positioning compressive
sampling filter 18 in-line with the backscattered light, filter 18
is positioned in-line with light source 2 (i.e., the local
oscillator). Because the local oscillator light being sent to mixer
22 is much more powerful than the backscatter light being received
from waveguide 8, the local oscillator (i.e., light source 2) acts
as an amplifier.
[0031] By positioning compressive sampling filter 18 in-line with
the local oscillator light and in front of mixer 22, instead of
in-line with the backscatter light, a similar effect is created to
that of FIG. 1, whereby the backscattered light was blocked.
Instead, in distributed optical sensing system 20, the local
oscillator light sent from source 2 is blocked by compressive
sampling filter 18 at times when the spatial regions of waveguide 8
to be blocked are arriving at receiver 24 (comprised of 24A and
24B). The unblocked local oscillator light is also referred herein
as "reference light." Although the local oscillator light is being
selectively blocked by compressive sampling filter 18, the ultimate
effect is that compressive sampling filter 18 is still utilized to
selectively block portions of the backscattered light, as will be
described further below.
[0032] During operation of optical sensing system 20, an optical
pulse is sent into the sensing waveguide 8 using light source 2 and
modulator 4, as previously discussed. As the backscatter light
returns to the receiver side of system 20, it is mixed with the
original laser light, called the local oscillator in a homodyne
demodulator (a heterodyne scheme is also possible in alternate
embodiments). The local oscillator light being passed by
compressive sampling filter 18 is referred to as reference light,
and is much more powerful than the backscatter light, providing a
gain to the signal. The reference light is mixed in two ways in the
90 degree optical hybrid mixers 22 positioned in-line with
waveguide 8. In the "I" portion of the optical hybrid light signal
(also referred to as "mixed light") sent to balanced receiver 24B,
the local oscillator/reference and backscatter light is mixed
normally. In the Q branch of the optical hybrid light signal (also
referred to as "mixed light") sent to balanced receiver 24A, the
local oscillator/reference light is phase shifted by 90 degrees
relative to the backscatter light signal.
[0033] By providing an I and Q interferometric signal, the optical
phase can be demodulated by arctan(Q/I). In a distributed vibration
or acoustic sensing system, as the waveguide is strained due to
vibration, the phase of the optical backscatter signal changes.
Detecting this phase change allows detecting vibration that is
impacting the waveguide. The time domain multiplexing methods works
here in system 20, as it does in the more general system. However,
in the case of system 20, compressive sampling filter 18 (e.g.,
high-speed optical shutter or switch) is placed in-line with the
local oscillator source 2. This has the advantage that any optical
attenuation induced by compressive sampling filter 18 is not
applied to the weak backscatter signal, but instead is applied to
the much stronger local oscillator 2. Typically, the local
oscillator power can be easily adjusted to as high of a level as
needed, whereas the backscatter light is very weak and cannot be
strengthened further. In fiber optic sensing system 20, when
compressive sampling filter 18 is closed, the local
oscillator/reference light does not mix with the backscatter light.
Due to the mechanism of the 90 degree optical hybrid and balanced
detector, any light signal that is not part of the mixing between
local oscillator 2 and the backscatter signal is automatically
subtracted out by the low bandwidth balanced detectors 24A,B, and
sent to corresponding low-speed ADCs 26A,B. Therefore, when local
oscillator 2 is blocked, effectively there is no interferometric
light signal being detected by optical receivers 24A,B. Since this
has an almost identical effect as if compressive sampling filter 18
was placed in-line with the backscatter light, compressive sampling
filter 18 of system 20 still in essence selectively blocks desired
portions of the backscattered light.
[0034] Accordingly, the previous illustrative embodiments utilize a
compressive sampling filter optically coupled in-line with the
backscatter signal or local oscillator to selectively remove the
backscatter signal of selected sections of the waveguide based on
the pseudo-random spatial randomization signal specified by
compressive sampling. An alternative approach, however, is to apply
the compressive sensing methodology to the interrogation signal
instead of the backscatter signal. In the previously described
illustrative systems, the pulse is equal to or less than the
desired spatial resolution. For example, to achieve a one meter
spatial resolution, the pulse width must be 10 ns (1 meter two-way)
or less in duration. In distributed fiber optic sensing systems,
there is an upper limit to the instantaneous optical power that can
be sent into the fiber before negative effects occur. Primarily,
the negative effects occur when exceeding the maximum power density
supported by the fiber. There are several possible non-linear
optical is processes that are undesirable in measurement systems,
which include: stimulated Raman scattering, stimulated Brillouin
scattering, self-phase modulation and modulation instability. These
non-linear effects set a ceiling on the optical power level of the
pulse that may be used.
[0035] However, they do not set a ceiling on the total optical
energy of the pulse, which is a product of the power and time
duration of the pulse. To send a pulse with higher optical energy,
but the same optical power, it is necessary to use a longer
duration optical pulse. As the total optical energy is increased,
the optical power of the backscatter signal, which contains the
information being measured, will increase in proportion. The
increased backscatter power will enhance the signal-to-noise ratio
of the measured parameter (e.g., temperature, vibration, etc.).
This has many advantages. In order to achieve a particular spatial
resolution using other interrogation methods, the pulse width must
be less than the spatial resolution desired. However, applying the
compressive sampling methods of the present disclosure to the pulse
will remove this limitation.
[0036] Therefore, in certain alternative embodiments of the present
disclosure, regions of the pulsed interrogation signal are blocked
using a spatially pseudo-random filter applied in the time domain.
FIG. 3 illustrates a distributed optical sensing system in which
compressive sampling is applied to the light pulse, according to
certain illustrative alternative embodiments of the present
disclosure. The embodiment of FIG. 3 is similar to previous
embodiments, where like numerals refer to like components. However,
in this embodiment, distributed optical sensing system 30 includes
a compressive sampling filter 18 is optically coupled between light
source 2 and waveguide 8 so that the interrogation signal (i.e.,
the pulsed light sent from modulator 4) is selectively blocked.
Here, compressive sampling filter 18 selectively blocks portion of
the light pulse 28 to thereby generate a compressive sampling pulse
32.
[0037] With reference to FIG. 3, for example, at a particular time
after pulse 28 is emitted from modulator 4, the backscatter light
reaching receiver 12 will consist of the region of waveguide 8 that
is at a distance with the two-way travel time from the interrogator
equal to the time interval after pulse 28 was emitted. It is clear
that at a particular moment of time after pulse 28 is emitted, the
backscatter from the front of pulse 28 will be originating from a
location on sensing waveguide 8 more distant than the backscatter
generated by the back of pulse 28. It works out, due to the two-way
travel time of the backscatter light, that the backscatter light at
the receiver 12 will be exactly a combination of spatial locations
that match the pattern within pulse 32 is generated by compressive
sampling filter 18, but with pulse 32 spatially compressed by a
factor of two due to the two-way travel nature of time-domain
reflectometry. Since this has an almost identical effect as if
compressive sampling filter 18 was placed in-line with the
backscatter light, compressive sampling filter 18 of system 30
still in essence selectively blocks desired portions of the
backscattered light.
[0038] Therefore, those regions of compressive sampling pulse 32
that have been blocked will represent spatial locations on
waveguide 8 that are not emitting backscatter at receiver 12 at a
particular moment in time, and those parts of pulse 32 that are not
blocked will produce backscatter that reaches optical receiver 12
at the same moment that the blocked regions are not providing
backscatter to optical receiver 12. Thus, during operation of
distributed optical sensing system 30, a series of compressive
sampling pulses 30A-C are sent down the fiber with a different
spatial time-domain filter applied to each pulse as shown in FIGS.
3 and 4. FIG. 4 is a diagram showing three different compressive
sampling pulses 32 at a particular moment overlaid on sensing
waveguide 8. The diagram includes dashed lines corresponding to the
regions of waveguide 8 which are providing backscatter to optical
receiver 12 (high levels of pulses 32A-C) and which regions (low
levels of 32A-C) are not. This is only for one instant in time. At
a moment later, for example 100 ns later, the backscatter light
being received at optical receiver 12 will be represented by the
pulse shifted ten meters further down waveguide 8.
[0039] Therefore, the compressive sampling methods described herein
can be implemented using an interrogation pulse with sections of
the pulse blocked. The spatial resolution achieved with this method
is based on the switching rate of the compressive sampling filter
(e.g., optical switch), not on the width of the pulse. The result
is that a very wide pulse (i.e., 1 microsecond or 10 microseconds,
or longer), and thus a very high total energy pulse, may be emitted
into the sensing waveguide. This pulse will provide a backscatter
signal of much higher power than a short pulse (e.g., 10 ns or 20
ns) which is traditionally required to maintain a high spatial
resolution of typically around one or two meters. By sending a
series of very wide, high energy pulses, each with a pseudo-random
time-domain filter applied to the pulse, the return backscatter
signal collected over many pulses can be "de-compressed" via
compressive sampling algorithms to achieve the original high
spatial resolution, but with the benefit of a much higher signal to
noise ratio provided by the higher backscatter power levels.
[0040] In conventional OTDR systems, after the pulse is launched,
the system cannot launch another pulse until the original pulse has
traversed to the end of the waveguide and all backscatter returns
generated by that pulse have reached the optical receiver. If a
second pulse was introduced before all backscatter had returned to
the interrogator, the optical receiver would effectively be
receiving backscatter from two locations on the fiber at the same
time, thus creating a cross-talk noise problem between the two
sensing locations. These two locations being received at the
optical receiver would slide down the fiber over time as the two
pulses traversed down the fiber.
[0041] However, with the illustrative compressive sensing methods
described herein, the cross-talk problem is overcome because the
locations on the waveguide that are being simultaneously received
can be spatially selected by a pseudo-random spatial pattern
provided by the compressive sensing spatial randomization signal.
This will allow the ability to sending multiple pulses into the
waveguide at once, with each pulse having a different pseudo-random
spatial pattern. Extending this concept further, in certain
illustrative embodiments there will be no separation in time
between the end of one pseudo-random pulse and start of the next.
The end effect is that pseudo-random modulated light is
continuously sent into the waveguide. In this scheme, the optical
backscatter power will be maximized and at any instant in time the
combination of spatial locations on the waveguide being received at
the optical receiver will be based on a pre-determined
pseudo-random spatial pattern.
[0042] In the illustrative embodiments described herein, processing
circuitry is communicably coupled to the sensing systems as
previously described. The signals recorded by the optical receiver
are decompressed using compressive sampling algorithms to recover
the parameter being sensed (e.g., temperature, vibration, etc.) at
each sensing location in the dense sensor array provided by the
fiber or other waveguide, but with much higher signal to noise
ratio than would be possible using conventional OTDR based sensing
schemes. Illustrative compressive sampling recovery algorithms to
decompress the acquired low bandwidth signal include, for example,
l.sub.1-norm minimization routines such as Orthogonal Matching
Pursuit ("OMP") and derivatives thereof such as Stagewise OMP and
Regularized OMP.
[0043] In alternative embodiments of the present disclosure, the
optical sensing systems described herein may include any number of
interrogation arms. For example, the backscattered light leaving
optical circulator 6 may be split into several interrogation paths
and process is independently by a compressive sampling filter
controlled by a unique spatial randomization signal. Thus, the
blocked portions of the backscattered light would vary between each
interrogation arm. These and other modifications of the present
disclosure will be apparent to those ordinarily skilled in the art
having the benefit of this disclosure.
[0044] Although the illustrative distributed optical sensing
systems described herein may be used in a variety of applications,
the following example will discuss a wellbore application. FIG. 5
is a schematic illustration of a distributed optical sensing system
extending along a wellbore, according to alternative illustrative
embodiments of the present disclosure. The waveguide may be
comprised of a single waveguide or an array of waveguides. In the
illustrated embodiment, however, distributed optical sensing system
50 includes waveguide detector 52 that is comprised of an array of
waveguides 52a-d. Distributed optical sensing system 50 includes a
light detection unit 54 positioned at a surface location 56. In
other embodiments, however, light detection unit 54 may be
positioned downhole. Light detection unit 54 is part of the signal
processing chain, and may be any of the interrogation and receiving
components of the compressive sensing system described herein.
[0045] Light detection unit 54 described herein may include at
least one processor, a non-transitory, computer-readable storage,
transceiver/network communication module, optional I/O devices, and
an optional display (e.g., user interface), all interconnected via
a system bus. The network communication module may be any type of
communication interface such as a fiber optic interface and may
communicate using a number of different communication protocols.
Software instructions executable by the processor for implementing
the compressive sensing described herein may be stored in suitable
storage or some other computer-readable medium.
[0046] Moreover, those skilled in the art will appreciate that the
disclosure may be practiced with a variety of computer-system
configurations, including hand-held devices, multiprocessor
systems, microprocessor-based or programmable-consumer electronics,
minicomputers, mainframe computers, and the like. Any number of
computer-systems and computer networks are acceptable for use with
the present disclosure. The disclosure may be practiced in
distributed-computing environments where tasks are performed by
remote-processing devices that are linked through a communications
network. In a distributed-computing environment, program modules
may be located in both local and remote computer-storage media
including memory storage devices. The present disclosure may
therefore, be implemented in connection with various hardware,
software or a combination thereof in a computer system or other
processing system.
[0047] Nevertheless, light detection unit 54 is optically coupled
to waveguide detector 52, which may be transparent optical fibers
that extend over a desired range without significant reduction in
light intensity. Waveguide detector 52 extends down a wellbore 58
which has been drilled in a formation 60. During operation of the
embodiment of FIG. 5, the interrogation signals are sent from light
detection unit 54, received and processed using the compressive
sampling methods described herein to determine one or more
parameters of the downhole environment (e.g., pressure,
temperature, etc.).
[0048] The wellbore application illustrated in FIG. 5 may take a
variety of forms. For example, the application may be an onshore
oil or gas drilling rig in which the distributed optical sensing
system has been permanently or temporarily positioned along the
wellbore. In certain embodiments, the waveguide may form part of a
casing string. In other embodiments, the waveguide may be embedded
within the cement used to secure the casing in place. In other
applications, the waveguide(s) may be deployed along a wireline or
other conveyance (e.g., logging-while-drilling or
measurement-while-drilling assembly, or some other suitable
downhole string). In yet other embodiments, the waveguide(s) may be
positioned inside the casing or may form part of an inner string,
such as, for example, part of a production string. There are a
variety of other ways in which these components may be permanently
or temporarily positioned downhole, as these are only illustrative
in nature.
[0049] Accordingly, the foregoing embodiments provide many
advantages. The disclosed compressive sampling optical systems can
be used to provide unprecedented spatial resolution (sub-centimeter
DTS, for example), or configured to provide very high
signal-to-noise ratio backscatter signaling while maintaining
typical spatial resolution of one or two meters. The illustrative
embodiments may be applied to any distributed fiber optic sensing
method that utilizes optical time-domain multiplexing methods.
[0050] Embodiments described herein further relate to any one or
more of the following paragraphs:
[0051] 1. A distributed optical sensing system, comprising a light
source to generate light; a waveguide positioned to optically
interact with the light to produce backscattered light; a
compressive sampling filter positioned to selectively block
portions of the light or the backscattered light based upon a
compressive sampling technique; and an optical receiver to receive
the backscattered light.
[0052] 2. A system as defined in paragraph 1, wherein the
compressive sampling filter is an optical switch or optical
shutter.
[0053] 3. A system as defined in paragraphs 1 or 2, wherein the
compressive sampling filter is optically coupled between the
waveguide and the optical receiver to selectively block portions of
the backscattered light.
[0054] 4. A system as defined in any of paragraphs 1-3, wherein the
optical receiver has a bandwidth of 10 MHz or less.
[0055] 5. A system as defined in any of paragraphs 1-4, wherein the
compressive sampling filter is optically coupled between a local
oscillator and a mixer to selectively block portions of the
backscattered light.
[0056] 6. A system as defined in any of paragraphs 1-5, wherein the
mixer is a 90 degree hybrid mixer.
[0057] 7. A system as defined in any of paragraphs 1-6, wherein the
optical receiver is a balanced optical receiver.
[0058] 8. A system as defined in any of paragraphs 1-7, wherein the
compressive sampling filter is optically coupled between the light
source and the waveguide to selectively block portions of the
light.
[0059] 9. A system as defined in any of paragraphs 1-8, wherein the
light source generates a light pulse having width of 1 microsecond
or longer.
[0060] 10. A system as defined in any of paragraphs 1-9, further
comprising processing circuitry communicably coupled to the
receiver to receive the backscattered light and thereby determine a
parameter being sensed by the waveguide.
[0061] 11. A system as defined in any of paragraphs 1-10, wherein
the waveguide is positioned along a wellbore.
[0062] 12. A system as defined in any of paragraphs 1-11, wherein
the waveguide is a fiber optic cable.
[0063] 13. A distributed optical sensing method, comprising
generating light that optically interacts with a waveguide to
produce backscattered light; selectively blocking portions of the
is light or the backscattered light using a compressive sampling
filter; and analyzing the backscattered light to thereby determine
a parameter being sensed by the waveguide.
[0064] 14. A method as defined in paragraph 13, wherein selectively
blocking the backscattered light comprises optically interacting
the backscattered light with the compressive sampling filter.
[0065] 15. A method as defined in paragraphs 13 or 14, wherein
selectively blocking the backscattered light comprises using an
optical switch or optical shutter to selectively block the
backscattered light.
[0066] 16. A method as defined in any of paragraphs 13-15, wherein
selectively blocking the backscattered light comprises optically
interacting a local oscillator light with the compressive sampling
filter to produce a reference light; optically interacting the
reference and backscattered light with a mixer to produce mixed
light; and optically interacting the mixed light with the receiver
to thereby generate the backscattered light.
[0067] 17. A method as defined in any of paragraphs 13-16, wherein
optically interacting the backscattered light further comprises
phase shifting the reference light 90 degrees relative to the
backscattered light.
[0068] 18. A method as defined in any of paragraphs 13-17, wherein
selectively blocking the backscattered light comprises optically
interacting the light with the compressive sampling filter to
produce a compressive sampling pulse having blocked regions.
[0069] 19. A method as defined in any of paragraphs 13-18, wherein
generating the light comprises generating a light having a pulse
width of 1 microsecond or longer.
[0070] 20. A method as defined in any of paragraphs 13-19, wherein
determining the parameter comprises determining a parameter along a
wellbore.
[0071] 21. A distributed optical sensing method, comprising
interrogating a sensing waveguide using an interrogation signal;
receiving backscattered signals from selected regions of the
sensing waveguide based upon a compressive sampling technique; and
using the backscattered signals, determining a parameter being
sensed by the sensing waveguide.
[0072] 22. A method as defined in paragraph 21, wherein receiving
backscattered signals comprises selectively blocking portions of
the backscattered signals using a compressive sampling filter
positioned in-line with the sensing waveguide.
[0073] 23. A method as defined in paragraphs 21 or 22, wherein
receiving backscattered signals comprises selectively blocking
portions of a local oscillator light.
[0074] 24. A method as defined in any of paragraphs 21-23, wherein
receiving backscattered signals comprises mixing the backscattered
light with a reference signal.
[0075] 25. A method as defined in any of paragraphs 21-24, wherein
receiving backscattered signals comprises selectively blocking
portions of the interrogation signal.
[0076] 26. A method as defined in any of paragraphs 21-25, wherein
a wellbore parameter is determined.
[0077] Although various embodiments and methodologies have been
shown and described, the disclosure is not limited to such
embodiments and methodologies and will be understood to include all
modifications and variations as would be apparent to one skilled in
the art. Therefore, it should be understood that the disclosure is
not intended to be limited to the particular forms disclosed.
Rather, the intention is to cover all modifications, equivalents
and alternatives falling within the spirit and scope of the
disclosure as defined by the appended claims.
* * * * *